Update app.py

app.py

@@ -3,8 +3,6 @@ import os, gc
 from datetime import datetime
 from transformers import CLIPImageProcessor
 from huggingface_hub import hf_hub_download
-from typing import List, Dict
-from dataclasses import dataclass
 DEFAULT_IMAGE_TOKEN = "<image>"
 
 
@@ -56,7 +54,6 @@ def generate(
     out_str = ''
     occurrence = {}
     state = None
-    print("in shape", model.w["emb.weight"].shape)
     for i in range(int(token_count)):
         if i == 0:
             input_ids = (image_ids + pipeline.encode(ctx))[-ctx_limit:]
@@ -105,10 +102,8 @@ def chatbot(image, question):
     image = image_processor(images=image.convert('RGB'), return_tensors='pt')['pixel_values']
     image_features = visual_encoder.encode_images(image.unsqueeze(0))
     emb_mixer.set_image_embeddings(image_features.squeeze(0))
-    print(emb_mixer.embedding.shape)
+    global model
     model.w["emb.weight"] = emb_mixer.get_input_embeddings()
-    print(emb_mixer.get_input_embeddings().shape)
-    print("out shape", model.w["emb.weight"].shape)
     image_ids = [i for i in range(emb_mixer.image_start_index, emb_mixer.image_start_index + len(image_features))]
     input_text = generate_prompt(question)
     for output in generate(input_text, image_ids):